Re: Broadband truth in advertising, redux

[Dave Farber <dave () farber net>]

Begin forwarded message:

> From: "David P. Reed" <dpreed () reed com>
> Date: February 10, 2010 11:00:32 AM EST
> To: dave () farber net
> Cc: ip <ip () v2 listbox com>
> Subject: Re: [IP] Re:    Broadband truth in advertising, redux

> There is no doubt that mobility (operation at 100 kph) creates a  
> design constraint that has significant costs - in fact the cost of  
> cell phones traveling at 1000 kph - airplanes is significant  
> disruption to the system, and is probably the *real* reason that not  
> using a cellphone on the airplane is a rule that the cellular  
> operators don't dispute.
>
> That said, cell systems are reasonably adaptive, and most phones are  
> not traveling, so the engineering of real cellular systems includes  
> design choices that minimize the costs for phones that are not  
> moving, since 90% of phones even in the US are not moving.
>
> It's not clear that the cost of supporting 10% of the phones moving  
> is really that high as a proportion of total opex.
>
> In a third world country that has a few cars, I would suggest that  
> mere tweaking of parameters in current designs is far better  
> engineering than designing a "one off" system that does not support  
> moving vehicles at all!   That's the Clean Slate philosophy, and it  
> is opposed to the basic "engineering pragmatics" that created the  
> Internet - interoperability and evolvability are extremely useful.
>
>
>
> On 02/10/2010 10:41 AM, David Farber wrote:
> > Begin forwarded message:
> >
> > From: Rahul Tongia <tongia.cmu () gmail com>
> > Date: February 10, 2010 9:12:51 AM EST
> > To: dave () farber net
> > Cc: ip <ip () v2 listbox com>
> > Subject: Re: [IP] Re: Broadband truth in advertising, redux
> >
> > Dave,
> >
> > While we're at the issue of fundamentals, I'd love for a  
> > quantification of the costs of *mobility* (hi speed vehicular soft  
> > handoff). One number I got from a respected source was 30%; another  
> > person said higher.  If we don't design everything for such, and  
> > developing countries could happily suffice with *portability* that  
> > should lower costs. More importantly, it could open the way for  
> > alternative technologies and designs than today.
> >
> > Thanks,
> >
> > Rahul
> >
> > On Wed, Feb 10, 2010 at 5:02 AM, David Farber <dave () farber net>  
> > wrote:
> >
> >
> > Begin forwarded message:
> >
> > From: "David P. Reed" <dpreed () reed com>
> > Date: February 9, 2010 6:20:11 PM EST
> > To: dave () farber net
> > Cc: ip <ip () v2 listbox com>
> > Subject: Re: [IP] Re: Broadband truth in advertising, redux
> >
> > I'm going to make one more try.   Can any of these experts please  
> > *quantify* these enormous costs incurred by cellphones that happen  
> > to be in the vicinity of a cell tower?  What *percentage* of  
> > capacity does a non-communicating cellphone occupy when it is not  
> > currently transmitting or receiving?
> >
> > In my *expert* opinion, these are insignificant, immaterial costs.   
> > And rightly so, since the cellular infrastructure is engineered to  
> > minimize such costs!
> >
> > This is essentially nonsense.
> >
> > On 02/09/2010 05:49 PM, Dave Farber wrote:
> > > Begin forwarded message:
> > >
> > > > From: Alex French <alex () evilal com>
> > > > Date: February 9, 2010 3:49:43 PM EST
> > > > To: dave () farber net
> > > >
> > > > Subject: Re: [IP] Re: Broadband truth in advertising, redux
> > >
> > > > > On 02/08/2010 02:14 PM, Brett Glass wrote:
> > > >
> > > > > > P.S. -- Oh, and Karl, lest I forget to mention it, any continuous
> > > >
> > > > > > connection between a cell site and a client incurs significant  
> > > > > > overhead.
> > > >
> > > > > > The cell has to reserve codes (CDMA), time slots (TDMA), or  
> > > > > > channels
> > > >
> > > > > > (OFDM) for any client with which it has established a  
> > > > > > conversation, and
> > > >
> > > > > > can't continuously do this for every cell phone in the vicinity.
> > > >
> > > > Dave,
> > > >
> > > > It's worth mentioning that this applies to wired broadband too;  
> > > > every
> > > > connection means a session allocated on an ISP BAS, and equipment  
> > > > is
> > > > usually spec-ed for X simultaneous PPPoE connections; in fact,  
> > > > Cisco
> > > > charge significant license fees to ISPs based on the number of
> > > > concurrent connections (active or not).
> > > >
> > > > --
> > > > Alex
> > > Archives        
> >
> > Archives        
> >
> > Archives        



-------------------------------------------
Archives: https://www.listbox.com/member/archive/247/=now
RSS Feed: https://www.listbox.com/member/archive/rss/247/
Powered by Listbox: http://www.listbox.com